CHAPTER 7 CONCLUSIONS
This chapter provides conclusions of the findings reported in the previous chapters. First, research questions are answered. Second, implications are addressed.
Finally, limitations of the present study and suggestions for future research are presented.
7.1 Main Findings of the Present Study
There were five research questions in the present study, whose answers will be addressed respectively in the following.
1. What are the similarities and differences in the performance of complaint behaviors produced by native speakers of American English and native speakers of Chinese?
In the investigation of the American and Chinese complaints, there were four pieces of evidence showing their similarities. First, the Americans and the Chinese tended to make complaints when facing an offense. Second, both groups of subjects demonstrated similar tendencies in terms of individual and two-strategy distributions.
They were found to use indirect complaint and request for repair most frequently and to use opting out and threat least frequently. The use of indirect accusation and direct complaint strategies were in between. Third, they tended to use off-record
strategies or patterns more frequently as opposed to on-record counterparts. Finally, they tended to use requestive complaints more frequently than non-requestive complaints.
Despite the similarities, there were pragmalinguistic and sociopragmatic differences found between the American and the Chinese performances.
Pragmalinguistic differences were observed in the linguistic forms of indirect accusation and request for repair. The Americans tended to use wh-questions to make
indirect accusations, while the Chinese tended to use yes/no questions with negative orientation. With regard to request for repair, the Americans tended to use hedged performative, while the Chinese tended to use want statement. On the other hand, the
major sociopragmatic difference was shown in the content of indirect complaint + request for repair. The Americans were found to state facts of annoyance regardless
of the interlocutor’s status. The way they showed politeness was to use external downgrading devices. However, the Chinese tended to be influenced by interlocutors’
status to a greater extent. For example, when facing a complaint situation with a higher-standing interlocutor, they tended to save the interlocutor’s face or to condemn the socially unacceptable behaviors.
2. To what extent does pragmatic transfer affect Chinese learners’ complaint behaviors in American English?
In ILP research, there has been a consensus that learners perform an intended speech act largely based on their L1 knowledge. Therefore, in the present study, the learners were predicted to conduct positive transfer if there were L1-L2 equivalences, and they were predicted to conduct negative transfer if there were L1-L2 differences.
However, it was found that transfer predictions did not always match transfer occurrences. In the quantitative analysis, positive transfer of on-record/off-record patterns and requestive/non-requestive complaints occurred, and negative transfer of indirect accusation and indirect accusation + indirect accusation was also observable
in learner performances. However, positive transfer predictions were partially confirmed in the use of indirect complaint, direct complaint, request for repair, on-record/off-record strategies and indirect complaint + request for repair.
The qualitative analysis also demonstrated that transfer predictions did not always correspond to transfer occurrences. With regard to request for repair, the learners negatively transferred want statement and positively transferred query
preparatory as expected. However, they did not use yes/no questions with negative
orientation to make indirect accusation. The content of indirect accusation + indirect complaint did not show L1 interference in the learner performances either. For
indirect complaint + request for repair, it was found that L1 sociocultural information was transferred to L2 in the high-proficiency learners’ performances, but such transfer was not observed in the low-proficiency learners’ performances.
For explanation, learners may have intended to transfer from L1, but confined by their linguistic proficiency, they may not have been able to conduct the transfer. In pragmalinguistic transfer, which refers to L1 form-function mapping in L2, the high-proficiency learners were found to conduct positive transfer more frequently than the low-proficiency learners, while the low-proficiency learners were found to conduct negative transfer more frequently than the high-proficiency learners. It was probably because the high-proficiency learners had more linguistic control over L2 to produce appropriate linguistic forms, but the low-proficiency learners did not have the capability to do so because of their lower level linguistic skills. On the other hand, the high-proficiency learners were more likely to conduct negative sociopragmatic transfer, which refers to L1 evaluation of social context to L2. This was probably because the high-proficiency learners had sufficient L2 knowledge to transfer L1 sociocultrual norms.
3. What are the similarities and differences between the Chinese learners’ productions and perceptions of the DCT and role plays?
The comparison between the learners’ DCT and role play performances revealed that they evidenced similar tendencies in strategy distribution. In both elicitation methods, the learners used indirect complaint and request for repair most frequently, then direct complaint and indirect accusation strategies, with opting out and threat used least frequently. However, the two methods differed in the number of
strategies within a single conversational turn. In the DCT, the learners tended to use two strategies to issue a complaint, while in the role plays, they tended to use only one strategy. It was probably because the DCT format required the learners to say everything important on the first turn. But in the role plays, the learners may not have had the opportunity to squeeze everything on turn one because in a natural conversation, the next speaker will take the floor at any transition relevance place.
The second difference between the learners’ DCT and role play performances was that the former was more grammatical than the latter. Since the DCT is a writing task, the learners had more time to monitor their performances. The role plays, however, are an oral task, so the learners did not have much time to plan what they would say in the flow of conversation. Finally, the content also reflected differences between the two methods. The DCT elicited the most important parts of an intended speech act, while the role play performances contained more conversational features in addition to head acts.
With regard to learners’ perceptions of the DCT and role plays, the learners reported that the common difficulty they faced was lexical deficiencies, but they worked on the problem differently within these two methods. Generally speaking, the learners tended to replace an intended word with another one in the DCT, but they tended to use communication strategies such as paralinguistic cues, repetitions and hesitations in the role plays. They also reported that the DCT was easier for them because they had total control over the task. The role plays were more challenging for the learners because they needed to give immediate responses. However, the learners liked role plays better because the role plays reflected more of real-life interactions, expressed emotions to a greater extent and were more interesting.
4. To what extent does explicit teaching affect the learners in the acquisition of complaint behaviors?
In the learners’ post-instructional DCT performances, the quantitative analysis showed that instruction seemed to work more effectively for the high-proficiency learners because they produced native-like performances with respect to overall strategy distributions, on-record/off-record strategies and requestive/non-requestive complaints. Instruction also worked for the low-proficiency learners because they made progress in the use of indirect complaint, direct complaint, indirect accusation, request for repair and on-record/off-record strategies, although their performances
were still not native-like. The qualitative analysis also showed that the learners made progress in the use of request for repair. The learners were found to decrease the use of mood derivable and want statement, and to increase the use of query preparatory.
In addition, they began to use past tense verbs/modals more frequently to mitigate requestive forces. However, there were certain aspects resistant to instruction in the learners’ performances after instruction. The first problem addressed the low frequencies of hedged performative. It is possible there might not have been enough input for the learners to acquire this request form during the instructional period. The second problem concerned the use of query preparatory variations. It was found that unlike the Americans, who used can/could and will/would pairs in a more balanced way, the learners still used can/could pairs with the same frequency after instruction as they did before instruction. For explanation, the first problem seems to come from teaching induced errors, while the second problem seems to be L1 interference.
5. What are the learners’ perceptions of explicit teaching of complaint behaviors?
To answer this question, the learners’ attitudes towards explicit instruction were investigated in terms of general impressions, teaching materials and learning difficulties. Most of the learners in the present study reported that they found the instruction helpful and useful because speech acts were not treated specifically in their previous courses. With regard to teaching materials, they mentioned that the textbook
provided them with lots of communicative activities and the handouts gave them specific rules to follow. They also suggested the use of video clips to facilitate learning. Finally, the learners felt that the most challenging part in instruction was that they needed to digest a lot of rules within a short period of time.
7.2 Implications
As can be seen from the above summary, the present study was designed to accomplish three major goals. First, it investigated the complaint behaviors conducted by the native speakers of American English and the native speakers of Chinese.
Second, it examined the extent of pragmatic transfer with respect to learners’
proficiency. It also attempted to explain learner deviations from pragmalinguistic and sociopragmatic differences between American English and Chinese. Finally, based on the learners’ productions, explicit instruction was designed to improve their complaint behaviors in American English, and the instructional efficacy was examined. In the following, both the theoretical and pedagogical implications are discussed.
7.2.1 Theoretical Perspectives
The investigation into American and Chinese complaints contributes to the debate on universality versus culture-specificity, which has been discussed extensively in pragmatics research. One line of research speaks for universality. It has been recognized that there are certain general mechanisms regulating human languages, which include conversational maxims (Grice, 1975), politeness theories (Brown & Levinson, 1987; Leech, 1983), and taxonomy of communicative acts (Searle, 1975). However, the other line of research disagrees with the claim for universality and maintains that speech acts vary in conceptualization and verbalization across languages since pragmatic knowledge is a reflection of cultural norms. The most representative proponent is Wierzbicka (1991), who argues that universality is actually Anglo-Saxon biased because modern pragmatics has been heavily influenced
by British and American philosophers.
In the speech act of complaining, it is found that universality and culture-specificity coexist. The languages investigated so far include Chinese, Danish, Korean, Hebrew, American English, British English and Canadian French (Hartley, 1998; Laforest, 2002; Murphy & Neu, 1996; Olshtain & Weinbach, 1987, 1993;
Trosborg, 1995; and the present study). The first universal is that when facing an offensive act, most competent adult members of a given society are assumed to make complaints rather than opt out of the situation. The speaker who opts out may be praised as being tactful and polite (Olshtain & Weinbach, 1993), but he/she also fails to achieve desired communicative goals. Although some of the complaint studies cited do not include opting out as one of the strategies as the present study does, this universal exists because it orients to the rationality of human beings. The second universal addresses the preference of strategies when making complaints. The populations investigated so far cluster their preferred strategies around the center of the directness scale which corroborates with Olshtain & Weinbach’s (1987: 203) claim that there is “a distinct bulging in the center and lessening of usage at the two extremes of the scale”. The last universal is that subjects across the above languages tend to make complaints off record rather than on record. Making complaints off record occurs when the complainer does not mention the complainee in the proposition. On the other hand, making complaints on record occurs when the complainer explicitly mentions the complainee in the proposition (Olshtain &
Weinbach, 1993). It has been found that except for native speakers of Hebrew, who make complaints on record most of the time, native speakers of Danish, Chinese, American English and British English tend to make complaints off record more frequently than on record.
Cross-cultural variations in complaints can be observed from the present study.
It is useful to adopt a particular taxonomy to describe cultural variations. Hofstede’s (1980, cited in Lustig & Koester, 1993) individualism-collectivism dimension, which describes the extent to which an individual relates to a larger group, is the most important property to distinguish one culture from another. Hofsteade (1980) created an individualism index (IDV) to assess a culture’s relative position in the individualism-collectivism continuum. It has been found that countries such as the United States, Australia, Great Britain, and Canada have very high IDVs. Conversely, countries such as Hong Kong, Singapore, Thailand, Taiwan and Peru have low IDV values. Cultures with a high IDV have highly individualistic populations. They place great importance on their personal rights and the autonomy of an individual is paramount. They take care of themselves and their immediate families only. The words which best reflect their attributes are “self”, “privacy”, and “independence”.
However, cultures with low IDV tend to be group-oriented. People living in those societies must be loyal to the group to which they belong. The absolute obligation to the group and the group’s best interest always come before individual interest, and the individual seeks to be taken care of by the group. Such cultural difference is reflected in American and Chinese complaints. As mentioned above, Americans value personal rights, so the majority tends to express facts of annoyance or dissatisfaction with less consideration of power and distance. On the other hand, Chinese value relative position in society, so they tend to vary their complaints with more consideration of the two social parameters.
In addition to the issue of universality and culture-specificity, the present study focuses on investigating the complaint behaviors performed by Chinese learners of English. The findings bear important pedagogical implications to pragmatics instruction, which are discussed as follows.
7.2.2 Pedagogical Perspectives
Most studies on speech act productions have measured learners’ proficiency either directly by using standardized language tests such as TOEFL, or indirectly by length of residence or study, and then matched learner performance with native speaker performance. The results of these studies have generally shown that learners’
language proficiency does not correspond to their pragmatic proficiency (Bardovi-Harlig & Dornyei, 1998; Kasper, 2001; Niezgoda & Rover, 2001; Schauer, 2006). As Bardovi-Harlig (1999: 686) argues, “high levels of grammatical competence do not guarantee concomitant high levels of pragmatic competence”.
According to Kasper (1996), there are three conditions for the acquisition of pragmatics to occur: existence of pertinent input, acknowledgment of the input, and lots of opportunities to achieve a high level of processing control to retrieve pragmatic knowledge effectively in various communicative contexts. Therefore, instruction plays a crucial role in the increase of learners’ pragmatic proficiency in an EFL environment since the classroom may be the only place which meets the three conditions.
The learners’ progress and their positive attitudes shown in the present study have demonstrated the efficacy of explicit instruction in an EFL language classroom.
In the design of pragmatics instruction, there are some principles which need to be taken into consideration. First, sociopragmatics should be taught prior to pragmalinguistics. This study has shown that when making complaints, part of the learner deviations were a reflection of different cultural norms. Therefore, sociopragmatics instruction serves as the best scaffolding to teach learners what counts as appropriate language use in a given speech community. Through the concepts of external/internal contexts and the comparisons between native and target cultures, learners may have a better understanding in why and how certain linguistic
forms of a given speech act behavior are realized in L2.
Second, the teaching flowchart shown on page 72 can be further expanded to include modificational patterns, which can be taught after social context and strategy realizations. Modifications can be divided into two major categories: internal and external. Internal modifications may serve to downgrade the force of a complaint, which include downtoners, understaters, hedges, subjectivizers, cajolers and appealers.
Or they may serve to upgrade the force of a complaint, which include intensifiers, commitment upgraders, and lexical intensification. External modifications are supportive moves which serve either to justify the speaker’s right to issue a complaint, or to provide face-saving arguments. These supportive moves may function at the structural level such as preparators, at the interpersonal level such as disarmers or at the content level such as providing evidence (Trosborg, 1995). The present study does not include modifications as the teaching target because it focuses on strategy use, but teachers are highly encouraged to show to learners how to modify their complaints based on the situation they encounter.
Third, teachers need to be aware of the fact that learners at different proficiency levels may not be influenced by L1 in a similar fashion. As shown in the present study, the high-proficiency learners tended to conduct negative sociopragmatic transfer from L1 to L2 because they had more language control. On the other hand, the low-proficiency learners tended to make pragmalinguistic errors because of their limited linguistic competence. Since the present study also shows that frequent input is essential to learner success, the L1-L2 sociocultural differences should be highlighted for more proficient learners, while accurate sociolinguistic forms should be emphasized for less proficient learners.
Fourth, as McLean (2004) argues, students often believe that there is only one strategy to perform a speech act and that this strategy works for all situations. The
learners’ self reports in the present study have also shown that they tend to use the same expression over and over again in different situations. One approach to fixing this problem is to increase learners’ strategy repertoire. In addition to individual strategies, teachers may show to learners how to combine strategies together since native speakers of English may use more than one strategy to perform speech acts such as apology, compliment, and refusal (cf. Beebe et al.,1990; Rose, 2001; Yu, 1999). The present study also shows that most of the Americans tended to combine two strategies to respond to a complaint situation. The other approach is to provide learners with natural, authentic discourse in which various social parameters are demonstrated. However, there is usually a gap between naturally-occurring and textbook conversations (Bardovi-Harlig et al., 1991; Boxer & Pickering, 1995; Grant
& Starks, 2001; Holmes, 1988; Scotton & Bernsten, 1988; Wong, 2002). It is therefore suggested that teachers use audio-visual information such as TV programs and movies, or refer to research findings to make their own handouts to compensate for lack of authenticity in current textbooks.
Finally, Cohen & Olshtain (1993) claim that learners with minimal planning tend to make pragmatic errors in their speech act performances. As shown in the present study, the learners reported that the DCT was easier for them and their productions indicated that they made fewer grammatical errors in this type of task. So DCTs seem to be a good tool for learners in the earlier stage of instruction to get familiar with the newly learned strategies and to allow them more time to plan and monitor what they intend to say. After learners have achieved mastery over the strategies, teachers can provide them with open-ended role plays. Output can facilitate language acquisition because the learner can notice the gap between what has been recognized and what has not been identified in L2 targets (Swain, 1985). It is also suggested that role play tasks, whenever possible, can be carried out with native
speakers, who can provide both positive and negative evidences of appropriateness in L2 (Long, 1996).
7.3 Limitations of the Study
The present study is subject to the following limitations. The first concerns the subjects. Since the subjects were college students, it is not certain whether the same results will be obtained from those of different age groups. Furthermore, the distinction between the low- and high-proficiency groups would have been larger if the subjects had passed different levels of GEPT (e.g. beginning and high-intermediate levels).
Second, for the convenience of research design, the social power and social distance variables had only two dimensions (+P/-P, +D/-D). However, human relationships are more than dichotomous. In fact, Olshtain & Weinbach (1987) argue that there are three types of social power: the speaker is higher than the hearer, the speaker and hearer are equal, and the speaker is lower than the hearer. There are also four types of social distance ranging from the least intimate to the most intimate:
strangers, acquaintances, friends and relatives. In addition, the inclusion of imposition would have added more value to the study because it could show how the Americans and Chinese perceive a given complaint situation.
Finally, although there seems to be some effects over the 25.5 hours of instruction in making the learners aware of more options to express complaints, it is not certain the extent to which instructional effects can be sustained. As shown in Koike & Pearson’s (2005) study, the explicit group performed significantly better than the control and implicit groups in the learning of Spanish suggestions and responses to suggestions, but the effects of such instruction and feedback was not represented in the delayed posttest. Therefore, it is questionable whether the instructional gains demonstrated in the present study can be retained in the longer term.
7.4 Suggestions for Future Research
The suggestions for future research are provided as follows. First, as mentioned in Chapter 2, there is no typical corresponding second part of a complaint because there is no conventional approach to recognize a complaint (Edmonson, 1981).
Further research can be devoted to the examination of complaint responses to shed light on complaint exchanges as a whole.
Second, in ILP research, some studies combined productions with metapragmatic assessment, whose purpose is to understand subjects’ perceptions of contextual variables in a given situation. Since American and Chinese cultures are different, it seems to be worthwhile to investigate whether Americans and Chinese hold different perceptions of external context (power and distance) and internal context (severity of offense and offender’s face loss) in given complaint situations.
Finally, except for level of proficiency and length of stay, other individual difference (ID) variables such as motivation and learner beliefs are seldom investigated in ILP research. It seems that the only study investigating learners’
motivation in pragmatics learning is Takahashi’s (2006) research, which investigated the relationship of pragmalinguistic awareness with proficiency and motivation. As a subset of SLA research, ILP researchers may find it interesting and challenging to explore this area of study.
